Civil Engineer in South Africa - Volume 26, Issue 10, October 1984

The Zuikerbosch Canal has attracted attention not only as a major feature in the board's programme of new works, but also because of the interesting and challenging technical problems that arose during its design, some of which are described in the papers in this issue of the institution journal.

Most canals with concrete linings have joints at regular close intervals to control cracking of the concrete due to shrinkage and thermal movements. Whereas such canals are usually taken out of service for regular maintenance of the joint sealants, the Zuikerbosch canal was designed to operate continuously. The concrete lining was therefore designed to have no joints at all, with the size and spacing of the cracks controlled by high tensile steel wire reinforcement in accordance with the then newly published British Code BS 5337 for water retaining structures.

---

Crack spacings and crack widths were found to be larger than the values calculated according to BS 5337. Nevertheless a concrete canal lining without joints was shown to be a workable proposition.

Kanale wat met beton uitgevoer is, het meesal voeë op kort reëlmatige afstande om krake wat weens krimping en termiese beweging ontstaan, te beheer. Terwyl sulke kanale gewoonlik buite bedryf gestel moet word vir die gereelde instandhouding van voegverseëling, is die Zuikerboschkanaal ontwerp om voltyds bedryf te word. Die voering van beton is gevolglik ontwerp om geheel en al sonder voeë te wees terwyl die wydte en spasiëring van krake beheer word deur draad wapening van trekvaste staal wat voorsien is volgens die toenmalige pasgepubliseerde Britse Kode BS 5337 vir waterhoudende strukture. Daar is gevind dat kraakwydtes en kraakspasiëring in werklikheid groter is as die waardes volgens BS 5337 bereken. Nogtans is dit getoon dat 'n betonkanaalvoering sonder voeë volkome uitvoerbaar is.

The Zuikerbosch canal was required to be a high security canal designed to operate continually without down-time for maintenance. It was therefore necessary to take steps to ensure that the highly expansive slickensided clays encountered along the canal route would not produce differential soil movements sufficient to damage the thin concrete lining of the canal.

---

The potential heave of the expansive clays was dealt with by prewetting along the entire length of the canal.

---

Slope stability analyses showed that the expansive slickensided clays would have to be over-excavated beyond the lines of the canal concrete lining and replaced with an adequately thick blanket of compacted clayey sand backfill excavated from carefully selected borrow pits. The expansive soils were identified and classified as being of either high or low expansiveness by means of the commonly used parameters clay fraction, liquid limit and plasticity index. The classification was used in conjunction with visual observations of the extent of slickensiding in the soil horizons.

---

The canal was commissioned in May 1983 and is operating satisfactorily.

Die Zuikerboschkanaal was bedoel om 'n hoogs veilige kanaal te wees, en is ontwerp om deurentyd in gebruik te wees sonder onderbrekings vir instandhoudingswerk. Dit was dus nodig om te verseker dat die hoogs swelbare wryfspieëlklei langs die kanaalvoete nie veranderlike grondverskuiwings veroorsaak wat die dun betonvoering van die kanaal kan beskadig nie.

---

Om potensiële uitswelling van die swelbare klei teen te gaan, is die hele afstand van die kanaal vooraf benat.

---

Ontledings van die stabiliteit van die skuinstes het aan die lig gebring dat die sweibare wryfspieëlklei weer uitgegraaf sou moes word tot anderkant die lyn van die kanaal se betonvoering. En dat dit vervang moes word met 'n dik genoeg en omvattende laag verdigte kleiagtige sand uit versigtig uitgesoekte steelgroewe. Die swelbare grond is deur middel van die gangbare parameters kleibreuke, vloeistofperke en plastisiteitsgetal geïdentifiseer en ingedeel in kategorieë van hoë of lae swelbaarheid. Die indeling het gepaard gegaan met visuële waarnemings van die omvang van die wryfspieël in die grondhorison.

---

Die kanaal is in Mei 1983 in gebruik geneem en die bediening daarvan is bevredig.

The 500 Mℓ terminal reservoir (forebay) at the end of the Zuikerbosch canal presented a number of challenging design problems. The forebay is utilized as balancing storage between the Zuikerbosch Canal and the Rand Water Board's large water treatment plants. The problems derived from the shallow water depth and from the embankment foundations which comprised collapsing aeolian sands overlying permeable and active alluvial clays.

---

The cost of lining the 20 ha of the storage forebay with concrete, bentonite or plastic lining would have been exorbitant so a puddled in-situ compacted clay layer was used to seal the floor.

---

The design of the inlet, outlet and spillway structures was complicated by the shallow water depth and the rapid rate of draw-down in the reservoir. The performance of the structures has been satisfactory.

Die ontwerp van die 500 Mℓ terminale reservoir aan die einde van die Zuikerbosch-kanaal het met 'n aantal uitdagende probleme gepaard gegaan. Die reservoir word vir balanserende opgaring tussen die Zuikerboschkanaal en die Randse Waterraad se groot watersuiweringswerke gebruik. Die probleme was weens vlak water en walfondamente op windgedeponeerde swigsand wat oor waterdeurlatende bewegende spoelklei geleë is.

---

Om die 20 ha voorbekken met beton, betontoniet of plastiek uit te voer, sou buitensporig duur wees en gevolglik is 'n in-situ gekompakteerde stopkleilaag gebruik om die vloer te verseël.

---

Die vlak water en die hoë dalingstempo van die reservoir se watervlak het die ontwerp van die inlaat, uitlaat- en oorloopstruktuur bemoeilik. Die strukture werk bevredigend.

The Zuikerbosch canal is a good example of the difficulties that are experienced in establishing the grass needed to counter soil erosion and to heal the scars left on the landscape by civil engineering activities.

---

Successful grassing depends on the choice of the right types of grass, combined with regular fertilizing and mowing during the growing season. Civil engineering structures are seldom designed with these maintenance procedures in mind, with the result that the procedures are often more difficult and costly than they need be.

Hoësnelheidwaterstrale ononderbroke losgelaat vanaf radiaalsluise in die dempkomme by die Zoekfonteinpypleiding-beheerwerke skep toestande wat kavitasie in die hand werk. Bronne van kavitasie is hoofsaakiik ruheid aan die betonmure en -vloere van die geute wat na die dempkomme lei en skuifsones tussen vinnig- en stadig bewegende water in die hidrouliese spronge in die dempkomme. Kavitasie vanaf die eerste bron is geminimaliseer deur die gebruik van spesiale bekisting om ultragladde betonoppervlakke te voorsien, maar hierdie maatreëls kan nie verwag word om geheel en al geslaag te wees nie en derhalwe is belugting ook gebruik.

---

Belugting van water losgelaat vanaf die radiaalsluise voorkom nie noodwendig kavitasie nie maar beskerm die betonoppervlakke teen die uitwerking daarvan. By Zoekfonteln word die waterstrale in lug omhul onmiddellik stroomafwaarts van die sluise om meesleuring moontlik te maak deur lug in die turbulente randlaag wat op die wateroppervlakke vorm, te meng.

High velocity water jets discharged continuously from radial gates into the stilling basins at the Zoekfonteln pipeline control works create conditions conducive to cavitation. Sources of cavitation are primarily roughtness on the concrete walls and floors of the chutes leading into the stilling basins, and shear zones between fast moving and slow moving water in the hydraulic jumps in the stilling basins. Cavitation from the first source was minimized by the use of special formwork to provide ultra-smooth concrete surfaces, but these measures could not be expected to be entirely successful, so aeration was also introduced.

---

Aeration of the water discharged from the radial gates does not necessarily prevent cavitation, but protects the concrete surfaces against its effects. At Zoekfontein the jets are enveloped in air immediately downstream of the gates to allow air entrainment by the mixing of air into the turbulent boundary layer that develops on the surfaces of the water.

The Zuikerbosch canal incorporates two cast in-situ concrete inverted siphonpipes, with a tear-drop cross section about 6 m high and lengths of 520 m and 390 m, respectively. Using a travelling form 12,5 m long, the contractor building the siphons needed a three day turn-around time in order to complete the work within a tight construction schedule. The programme called for the inner form to be stripped 12 hours after casting the concrete, and the outer form to be stripped after 36 hours.

---

Whether or not the concrete could attain the minimum strengths specified by the designers soon enough to permit the forms to be stripped depended on the temperature of the concrete. The night-time winter temperatures on the Highveld were known to be low enough to inhibit the development of concrete strength during the critical first 12 hours after the concrete was cast.

---

It was known that there is a relationship between the early strength of concrete and its age and curing temperature. To establish the relationship a laboratory test programme was set up which led to the development of a simple control curve for determining formwork stripping time, based on the monitoring of concrete temperature.

Die Zuikerboschkanaal sluit twee omgekeerde duikpype in van in-situ gegote beton met 'n peervormige deursnit, omtrent 6 m hoog en met lengtes van onderskeidelik 520 en 390 m. Met gebruik van 12,5 m lange verskuifbare bekistingstukke het die aannemer wat die duikpype gebou het 'n driedaagse wenteltyd nodig gehad om die werk binne 'n kort konstruksieperiode te voltooi. Die program het vereis dat die binneste bekisting 12 uur nadat die beton gegiet is, gestroop moes word en die buitenste bekisting na 36 uur.

---

Of die beton die minimum toelaatbare strektes sou bereik wat die ontwerpers gespesifiseer het vir die stroping van bekistings, was afhanklik van die temperatuur van die beton. Dit was bekend dat die hoëveldse nagtelike temperature gedurende die winter so laag is dat dit die aanwas van betonsterkte kan strem gedurende die kritieke eerste 12 uur nadat die beton gegiet is.

---

Dit was bekend dat daar 'n verhouding is tussen betonsterkte en die temperature en duur van nabehandeling. Om die verhouding vas te stel, is 'n laboratoriumtoetsprogram opgestel. Die resultate het daartoe gelei dat 'n eenvoudige beheerkromme ontwikkel is om die tyd vir stroping van bekisting te bepaal na geland van die temperatuur van die beton.

The 20 km long raw water canal between Vaal Dam and Zuikerbosch discharges into a large terminal reservoir which acts as a forebay for the Rand Water Board treatment plant and high-lift pumping station. A spillway was required for the reservoir to avoid overtopping of the embankments following a trip-out of the pumping station due to power failure. As an alternative to the conventional black-water siphon with its large discharges and hunting propensities, an air-regulated, smooth-discharge, white-water siphon spillway was chosen. The paper is concerned with its design and model-testing.

---

Planning and model-testing revealed a number of design problems. Care was needed in the design of the de-aeration mechanism in the siphon barrel and the downstream stilling basin to ensure that the siphon primed properly. On the upstream side the elevation and shape of the lip of the siphon had to be such that the air flow into the siphon was uniform to avoid hunting.The uniform inflow of air was upset by the draw-down of the water surface associated with the increase in velocity head at the siphon entrance, leading to instability in this region. Extensive modeling was needed to remedy the problem.

---

Full-scale tests of the siphon spillway at Zuikerbosch showed that although the de-aeration mechanism and stilling basin operated satisfactorily, the problem of upstream surging had not been fully resolved.

Die 20 km lang rouwater kanaal tussen Vaaldam en Zuikerbosch vloei uit in 'n groot eindreservoir wat as voorbak dien vir die Randwaterraad se suiweringswerke en hoë-druk pompstasles. 'n Oorloop is vir die reservoir benodig om oorstroming van die kanaalwalle te verhoed indien die pompe afskakel as gevolg van 'n kragonderbreking. 'n Lug-gereguleerde wit-water heweloorloop is gekies as 'n alternatief tot die konvensionele swart-water hewel met sy neiging tot groot uitvloei en swewing. Hierdie referaat behandel die ontwerp en modeltoetsing van eersgenoemde metode.

---

Die beplanning en modeltoetsing het 'n aantal ontwerpprobleme aan die lig gebring. Aandag moes geskenk word aan die ontwerp van die ontlugtingsmeganisme in die hewelbuis en die stroomaf stilbak om te verseker dat die hewel voldoende voorgevoer word. Die hoogte en vorm van die hewellip aan die stroomopkant moes sodanig wees dat die lugvloei eenvormig in die hewel is om swewing te verhoed. Die eenvormige invloei van lug is versteur deur die aftrek van die watervlak wat gepaard gaan met die toename in snelheidsdrukhoogte by die hewelingang wat lei tot onstabiliteit in hierdie gebied. Uitgebrilde modeltoetse was nodig om die probleem te oorbrug.

---

Volskaalse toets van die heweloorloop by Zuikerbosch het getoon dat alhoewel die ontlugtingsmeganisme en stilbak bevredigend funksioneer, is die probleem met stroomop stuwing nog nie bevredigend opgelos nie.